CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (http://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [GAMBIT] Periodic face meshes (http://www.cfd-online.com/Forums/ansys-meshing/98784-periodic-face-meshes.html)

 swillems March 19, 2012 06:43

Periodic face meshes

2 Attachment(s)
I have been trying to make face meshes for periodic boundary conditions. On the faces I want to link are multiple edges (see attachment non-linked.jpg). When I want to link these faces (left face with right face in the example) in Gambit I have to assign one vertice per edge (in this case one for the circle and one for the surrounding square). However, the location of the vertices on the spheres are different and I am unable to rotate them. When I link them this way (with the option Periodic on and reverse direction off) the problem showing in attachment linked.jpg occurs. The mesh arround the circle is wrong. I hope I have described the problem clear enough and that anyone has a solution for me. If any more information is needed, please say so.

ps. the circles are part of a half sphere body.

 -mAx- March 19, 2012 06:48

choose the axis-origin for links surface, and the next vertice along x-axis for the right surface
Also, since your both surfaces are not identical you may split your circles with point (U-value 0.5)
Gambit will add one more point on each circle. Now both surfaces should be exactly the same

 swillems March 19, 2012 07:02

1 Attachment(s)
When I select just the two vertices you suggested, if I understand them correctly, I get an error: "cannot link faces face.1, face.11 due to all loops of edges not being specified". When I add the vertices on the circles I get the same problem. In the attachment I showed which vertices I selected.

Maybe it is important to mention that the square face consists of 4 lines. The geometry is a parasolid import file from Unigraphics.

 -mAx- March 19, 2012 07:07

First split each circle (location / u-value 0.5)
Then choose your 2 vertices (one per surface)

 swillems March 19, 2012 07:50

Yes, that was it. I thougth I had tried that already, but I must have done something wrong. Thanks for your help!

 All times are GMT -4. The time now is 22:06.